In order to support the generation of large-scale statistics and/or metrics describing characteristics and/or preferences of a user community that it serves, an entity such as a cloud service provider may maintain a context information database. The context information database may be populated with various types of information collected from various client devices and their users. In order to respect user privacy considerations, the data in the context information database should be structured in such a way that “at large” database consumers cannot identify the users to which the various data values correspond, and cannot determine which data values may correspond or be related to each other. However, it may also be desirable that the database be structured in such a way that database consumers may receive user authorization to view particular personal information in the database.